This has been the most exciting month of my career.
The software framework I've been building for the better part of a decade just hit v0.1. The early results are beyond what I dared to project. Over 50 business intelligence sessions this month. Half a dozen major enterprise brands. A century-old global health NGO. Four different healthcare subcategories, from fashion brands to scrubs.
And somewhere in that sprint, a brilliant writing teacher and author gave feedback on the report writer's voice. It shook something loose. A workshop I gave in 2018 came flooding back — a fictional relationship therapist and life coach, designed as a prototype for AI agents that could actually hold space for human complexity.
I've been working on and with AI agents since 2004, first at NYU Gallatin studying complex adaptive systems, then at NYU ITP building interactive telecommunications projects, as part of what's now a two-decade research thread on collaborative sensemaking, collective intelligence, and cybernetics.
The Template That Couldn't Scale
Before any of this was software, it was field work.
Erin and I started Rock 'n Renew as a vehicle for something we could see clearly from the inside: local systems could be dramatically improved through targeted supplemental programming. Traditional science-based curriculum in schools, combined with hands-on training in permaculture design, biodynamics, ecological restoration, stormwater management, native plant installations, rain gardens, engineered wetlands.
The model put artists and musicians at the center of regional organizations. Touring artists became continuous power sources for legislation, local activism, public awareness campaigns, collective action. We worked with 350.org on the PowerShift campaigns and campus climate challenges. We partnered with NRDC, produced events with the Download Festival, Snoop Dogg, Modest Mouse, ran plastic bag ban campaigns.
The gaps were obvious. Working through the various layers of each system, we sought to perfect a model we could then scale through a kind of template that included all the administration and governance.
The tools were not there.
I tried, many times over 15 years, and found the same bottleneck. The points in the system where the resource requirements and administrative work simply didn't scale. Too much work for the artists. Too much reliance on trust and not enough oversight. Successful programs dependent on individuals who just made it work but couldn't train their replacement or hand off the reins when they were ready to move on.
Templates couldn't hold the amount of human innovation, intuition, social networking and regular networking. Network effects were dependent on fragile, sensitive, and distinct relationships that had no parallel elsewhere.
The Matchmaking Problem
Matchmaking between stakeholders — to share the massive amount of local knowledge that accrues with each successful project, event, or campaign — was nowhere close to accurately capturing the value flows between stakeholders. The conditional logic which acts as circuitry between value flow channels — marketing, advertising, public will, intent — was invisible to the systems that needed it most.
The raw signal log of transactions which represent the moving parts of a system, as it broadcasts the ongoing enactment of its world model, includes dynamic conditional logic. If/and, OR statements, graphs of arrays of associative trails. These signals are barely registering by the observers and network participants who attempt to search for examples of successful circuit design for their particular problem.
They have no reliable database to search. No consistent query language. No search terms they can wield in repeatable, programmatic fashion.
Instead, people scroll Reddit forums, clicking "next" through pages of results, searching for the needle-in-a-haystack information that is actually buried in the community they participate in every day. Occasionally, a rare keeper-of-keys person, a networking wizard, the unicorn known as the spoke-maker, comes along. They just know everyone you should know when you have this kind of problem.
You mention that you're preparing to hire a wetland engineer for a groundwater sampling project on the local river. They remember that the town next door did a study of that river last year. It has the information you need. You don't need to spend any money to get it.
The Wireframe Sessions
Seeing this problem play out across dozens of perspectives and industries over the last decade led to many wireframe sessions with David Muller, where we had a strong intuition about what the solution would look like.
There's seeing it first in your mind's eye. Then you scribble it on a notepad. Then you mock it up in a design program. Get a wireframe going. Even get the backend to populate data.
But the problem was never at that level.
We've had the technical systems to solve this at the database, server, and front-end design level for decades. The problem has always been at the individual knowledge worker level. The day-to-day notetaking. The emailing. The Slack and Telegram and Discord server-hopping.
- information capture.
- information storage.
- information retrieval.
- information traversal.
- querying and search.
- networking across social networks and identities.
- Identity management. Preference and intent management. Declarative logic. Coordination. Sensemaking. Collective intelligence.
What the Crypto Boom Left Behind
Crypto and blockchain opened a lot of doors into the viable problem space.
Some of those doors led off a cliff.
A lot of poisonous substances came from the crypto boom. But some trace amounts of gold and silver have been accruing amid the noise.
Self-sovereignty. Cryptographically guaranteed provenance. Intent. A ledger of all transactions. Finality.
All of these can be invisible. No tokens required to achieve the goals of the original vision. Token economics are supplemental — they provide affordances for a diverse stream of interests. We can satisfy those interests without any tradeoffs with conventional systems.
No visible trace of crypto need appear for those not interested. We don't elaborate as to whether Uber's app is written in TypeScript or Ruby on Rails. There's no reason to highlight that Totem Protocol provides the means for fully decentralized operation via onchain-first network routing.
It's a function of menu settings and configuration at the project level for any project using Totem Protocol's schema.
v0.1
ShurIQ. Totem Protocol. Perceptagon Interactive Design.
Fifty sessions this month. Major enterprise brands, a century-old NGO, four healthcare verticals, and the feedback from every one of them confirming what we drew on napkins a decade ago: the system works.
Not because the technology is new. Because the architecture finally matches the problem.
The sprint continues. Framework v0.1 ships by July 4. The content pipeline that produces this post is itself running on the system it describes. If that sounds circular, you're getting the idea.
More to come.
Jonny Dubowsky is co-founder of Sense Collective and architect of Totem Protocol. He has been designing multi-agent systems and collective intelligence frameworks since 2004.